MANET Loss Tomography Based on Circle-Movement Mobile Model

摘要

Mobile Ad hoc network (MANET) loss tomography is a process for inferring internal wireless links loss rates based on End-to-End network measurement. The crucial step in MANET loss tomography is topology identification. Since the simulation results based on existing mobile models may not correctly reflect the true MANET performance, a novel circle-movement mobility model (CMM) and its topology inference algorithm are presented. Although MANET topology is dynamic during simulation, it is relatively stable during each identifiable time. MANET loss tomography, which consists of unicast back-to-back packets probes sent from a sender to pairs of receivers, and the inference algorithms of links loss rates are proposed. NS2 simulations validate the presented model and inference algorithms.